Cladistic Biogeography of Plants in Australia and New Guinea: Congruent Pattern Reveals Two Endemic Tropical Tracks

Abstract

?Although congruence among multiple taxa is accepted as the basis of the cladistic approach to biogeography, seldom has congruence been tested for patterns within continents. Moreover, contiguous land areas are subject to confounding factors that might lead to incongru? ence, and areas of endemism are difficult to define. Most previous studies have analyzed single taxa and therefore could not test for congruence. Here, we report the first cladistic biogeographic analysis of multiple plant taxa (ll angiosperm groups) within continental Australasia. We have used Page's (1994, Syst. Biol. 43:58-77) new mapping method implemented in his program COM? PONENT 2.0, minimizing the optimality criterion leaves added, which is the number of terminal branches that need to be added to reconcile all input cladograms with the general area cladogram. A randomization test showed significant congruence, and a jackknife test indicated which parts of the general area pattern are robust. Conflicting patterns shown by widespread species may have been influenced by recant range expansions. At least two independent histories are postu? lated for taxa endemic to the tropics. One appears as a close relationship between the monsoon tropical areas, including southern New Guinea, and the other is a track (a group of areas with a common history) of successive differentiation along the east and south coasts. The Australian wet tropics (the Atherton area), although geographically proximate to the monsoon tropics, are part of the east coast track. Tasmania shows a very early vicariance from the rest of the continent, and the central arid areas appear related to adjacent coastal areas rather than to each other (contra Cracraft, 1991, Aust. Syst. Bot. 4:211-227). The postulated exchange of taxa with southeast Asia following late Miocene contact was not tested because of a lack of suitable taxon cladograms. [Cladistic biogeography; congruence; tree mapping; Australia; New Guinea; angiosperms; trop? ics.1 Congruence is the essence of cladistic biogeography. If general area patterns ex? ist, reflecting a common history among multiple, distantly related taxa, then at? tempts to reconstruct that history are jus? tified. However, congruence cannot be as? sumed a priori; different taxa may respond differently to earth events, depending upon their varying vagility or response to the appearance of barriers to dispersal. A bar? rier to one species may be a suitable hab? itat for another. For example, black soil plains would be a barrier to both the plant genus Persoonia and the bird genus Platycercus, whereas calcareous plains support? ing forest would be a barrier only to Per? soonia. Thus, a vicariance event for one taxon may not be a vicariance event for an4 E-mail: mike.crisp@anu.edu.au. other, depending upon their differing eco? logical tolerances. Neither can congruence be expected for areas having multiple his? tories involving different taxa. Even if con? gruence is demonstrated, no single causal explanation, e.g., vicariance, can be as? sumed (Page, 1988). Nonetheless, congruent patterns have been demonstrated among taxa occurring on different continents, at least in the southern hemisphere (Brundin, 1965; Ro? sen, 1978; Edmunds, 1981; Humphries and Parenti, 1986; Weston and Crisp, 1994; Un? der and Crisp, 1996), and these patterns have usually been explained by the plate tectonic theory of the breakup of the supercontinent Gondwana during the last 160 million years, causing vicariance of widespread ancestral species. However, con? gruence at this level is sometimes viewed as trivial; the present continents are sepa-